Overview of Capitalism and Socialism for Effective Altruism

(mod­er­ate up­dates since origi­nal post, in keep­ing with CSS up­dates)

This is an ed­ited ex­cerpt from the lat­est draft of the Can­di­date Scor­ing Sys­tem. I figured it would be use­ful for peo­ple who are in­ter­ested in cap­i­tal­ism and so­cial­ism. It’s far from perfect, but I have not found any re­motely satis­fac­tory overview any­where else. This post can be a start­ing point for peo­ple who want to de­bate whether eco­nomic sys­tems should be treated as a pri­or­ity in EA.

Cap­i­tal­ism is an eco­nomic sys­tem where the means of pro­duc­tion—land, ma­chin­ery, in­vest­ment cap­i­tal—are mostly con­trol­led by pri­vate own­ers. So­cial­ism is an eco­nomic sys­tem where they are con­trol­led by the pub­lic in a col­lec­tive or gov­ern­men­tal or­ga­ni­za­tion. For a va­ri­ety of rea­sons, it could be im­por­tant to take a stance on the ques­tion of which is prefer­able.

Ba­sic concepts

How­ever, it’s not straight­for­ward to judge cap­i­tal­ism and so­cial­ism be­cause the ques­tion is sim­ply too vague to be a valid ba­sis for di­rect re­search. There are many very differ­ent ways that these sys­tems can be re­al­ized, so mod­ern eco­nomic and poli­ti­cal sci­ence schol­ar­ship fo­cuses on nar­rower and bet­ter-defined is­sues. Ace­moglu and Robin­son (2015) note, “we do not be­lieve the term cap­i­tal­ism to be a use­ful one for the pur­poses of com­par­a­tive eco­nomic or poli­ti­cal anal­y­sis… both Uzbek­istan and mod­ern Switzer­land have pri­vate own­er­ship of cap­i­tal, but these so­cieties have lit­tle in com­mon in terms of pros­per­ity and in­equal­ity be­cause the na­ture of their eco­nomic and poli­ti­cal in­sti­tu­tions differs so sharply.” Dis­course about so­cial­ism is there­fore a lot like the main­stream so­ciolog­i­cal con­cep­tion of race: the dis­crete cat­e­go­riza­tion is a very poor model for ac­tu­ally un­der­stand­ing things, but be­cause so many peo­ple and in­sti­tu­tions have wed­ded them­selves to con­structed cat­e­gories, we are forced to grap­ple with them.

Be­cause of this vague­ness, it is tough to pin down so­cial­ism in a way that is easy to judge. Deep ide­olog­i­cal dis­putes ex­ist among leftists in Amer­ica, con­crete policy pro­pos­als are rare and con­tro­ver­sial, and there is no clear con­cep­tion of what so­cial­ism would ac­tu­ally look like (New Repub­lic). The philoso­pher Slavoj Žižek (book 2011), in a gen­eral opus of leftist anal­y­sis and re­sponse to the 21st cen­tury world, did not in­clude a clear idea of a path for­ward, ul­ti­mately con­clud­ing “we do not know what we have to do, but we have to act now, be­cause the con­se­quences of in­ac­tion could be catas­trophic.” The So­cial­ism 2019 con­fer­ence con­tained al­most no dis­cus­sion of how to struc­ture a so­cial­ist econ­omy.

So if we want to eval­u­ate the de­sir­a­bil­ity of generic in­creases or de­creases in the prob­a­bil­ity of so­cial­ist change, we can­not se­lect a spe­cific policy pro­posal that we would pre­fer. In­stead we must sur­vey the var­i­ous pos­si­ble forms and com­po­nents of so­cial­ism and pro­duce a vague ex­pec­ta­tion over the lot­tery of pos­si­bil­ities.

Note that Karl Marx be­lieved that so­cial­ism (fol­lowed by com­mu­nism) will in­evitably re­place cap­i­tal­ism sooner or later, in which case this judg­ment would re­quire a slightly differ­ent fram­ing about the timing that we pre­fer for the tran­si­tion. How­ever, Marx’s the­ory of his­tor­i­cal ma­te­ri­al­ism doesn’t ac­tu­ally show that cap­i­tal­ism will end or that its suc­ces­sor would be com­mu­nism (Jonathan Wolff book 2003, p. 111). Ad­di­tion­ally, his­tor­i­cal events have shown a re­verse phe­nomenon, that so­cial­ism served as a tran­si­tion state from feu­dal­ism to cap­i­tal­ism (Milanovic book 2019, Quillette re­view).

The pic­ture from his­tor­i­cal and eco­nomic evidence

First, when it comes to the cen­trally planned eco­nomic mod­els of 20th cen­tury so­cial­ist states, economists over­whelm­ingly re­gard them as in­fe­rior to cap­i­tal­ist economies. This con­sen­sus is sup­ported by the rele­vant pub­lished re­search. There is dis­agree­ment on whether early Soviet in­dus­tri­al­iza­tion was ex­pe­d­ited much by cen­tral plan­ning (Krug­man 1994, Allen 2005), but it had tremen­dous hu­man costs nonethe­less. The en­tire suite of so­cial­ist poli­cies re­duced Soviet agri­cul­tural pro­duc­tivity by about 50% (John­son and Brooks 1983). Cen­tral plan­ning went awry in the USSR when poor lead­er­ship ap­peared later on (Allen 2001, Allen 2005).

East Ger­many’s eco­nomic sys­tem failed (book 2010).

The com­mu­nist sys­tem has been dis­as­trous for Cuba, the US em­bargo is not the sole cause of their prob­lems (Jales et al 2018, Salazar-Car­rillo and No­darse-León 2015, Ribeiro et al 2013, Ward and Dev­ereux 2012). Cuba was also buoyed by sig­nifi­cant Soviet aid and trade sub­sidies dur­ing the Cold War, and more re­cently is sup­ported by $5 billion in an­nual re­mit­tances from ex­pa­tri­ates, a very large amount rel­a­tive to the size of its econ­omy which might give it a bonus suffi­cient to greatly ex­ceed the losses from the em­bargo. We haven’t been able to find any re­li­able in­for­ma­tion on the mag­ni­tude of losses caused by the em­bargo, but it’s prob­a­bly not very large be­cause it does not re­strict Cuba’s trade with any other coun­try be­sides Amer­ica and it still al­lows a few kinds of trade with Amer­ica. While Cuba’s liter­acy and health­care met­rics are re­port­edly very high now, they were also rel­a­tively high be­fore the rev­olu­tion and have not im­proved at a stel­lar rate (Salazar-Car­rillo and No­darse-León 2015). Also, Cuba’s ac­tual health­care qual­ity is worse than re­ported (Ber­dine et al 2018). A minor bright spot is that their hur­ri­cane pre­pared­ness is quite good (Ja­cobin).

Not only is there a con­sis­tent em­piri­cal trend, but there are plau­si­ble the­o­ret­i­cal ex­pla­na­tions for eco­nomic failures in cen­trally planned so­cial­ism. One is the fa­mous “calcu­la­tion prob­lem,” an­other is the loss of ap­pro­pri­ate in­cen­tives, an­other is Sh­leifer and Vishny (1991)’s ar­gu­ment that cen­tral plan­ning cre­ates op­por­tu­ni­ties for plan­ners to ar­tifi­cially cre­ate short­ages in or­der to col­lect bribes.

A re­cent defense of cen­tral plan­ning is pre­sented by Philips and Roz­worski (book 2019), who ar­gue that large cor­po­ra­tions demon­strate the vi­a­bil­ity of cen­tral plan­ning in the con­text of mod­ern age. Nei­ther au­thor is a pro­fes­sional economist and at a glance we’re not sure if this ar­gu­ment would work: the prob­lem with cen­tral plan­ning is not merely the idea that it is too com­plex a task, but that cred­ible price sig­nals are nec­es­sary and that mem­bers of the gov­ern­ment will have bad in­cen­tives. Also, if effec­tive cen­tral plan­ning were pos­si­ble, we should ex­pect to see coun­tries like Cuba or China do­ing it by now. How­ever we should not dis­miss the book with­out read­ing it or see­ing cred­ible eco­nomic re­views. Un­til we can in­ves­ti­gate the mat­ter fur­ther, we can al­low a pos­si­bil­ity that effec­tive cen­tral plan­ning might be pos­si­ble.

And one could ar­gue that a demo­cratic com­mand econ­omy would perform bet­ter. Democ­racy does not di­rectly pro­mote eco­nomic growth, but it does have pos­i­tive in­di­rect effects (meta-anal­y­sis 2007). This seems like it would par­tially ex­plain the eco­nomic failures of 20th cen­tury so­cial­ism, but not fully.

But it may be the case that com­mand economies ac­tu­ally lead to au­toc­racy, which in turn usu­ally im­plies non­democ­racy. In that case they not only lack this ex­cuse for their eco­nomic failures but are also re­spon­si­ble for some of the ter­rible re­pres­sions com­mit­ted by their to­tal­i­tar­ian gov­ern­ments, in­clud­ing the Cam­bo­dian Kh­mer Rouge regime which was prob­a­bly the most mur­der­ous regime in mod­ern his­tory on a per-cap­ita ba­sis. Every sin­gle so­cial­ist regime has been au­thor­i­tar­ian, ei­ther from the out­set or as an even­tual out­come (al­though oth­ers were not as ex­treme as Stal­inist Rus­sia, Maoist China or the Kh­mer Rouge). This was of­ten in spite of nom­i­nally good in­ten­tions. Stalin be­lieved in the power of democ­racy and in us­ing per­sua­sion rather than mil­i­tary force (Stalin 1921), and the Soviet Union was sup­posed to de­rive its power from demo­crat­i­cally elected work­ers’ coun­cils (“so­viets”); this later turned out to be mean­ingless. East Ger­many had a par­li­a­ment, but it lacked real power to change eco­nomic plans (eco­nomic his­to­rian book).

One ex­pla­na­tion for the trend of so­cial­ist au­toc­racy is that cen­tral plan­ning in­volves a huge ar­ray of trade­offs and de­ci­sions to make, so only a bu­reau­cratic elite can meet its bur­dens and they must en­sure that ev­ery­one fol­lows the plans (think tank book, pp. 44-49). It im­plies that any demo­cratic sys­tem for cen­tral plan­ning would have to dis­pense with ro­bust checks and bal­ances, em­pow­er­ing a leader or a se­lect few with sweep­ing ex­ec­u­tive au­thor­ity over mat­ters of gov­ern­ment. Another worry is that cen­tral plan­ning of the econ­omy cre­ates an enor­mous con­cen­tra­tion of power in the gov­ern­ment, lay­ing the foun­da­tions for to­tal­i­tar­i­anism (think tank book, pp. 49-52). The Marx­ist philoso­pher Slavoj Žižek (2011) ac­knowl­edges that “there is a grain of truth” in Ayn Rand’s idea that re­mov­ing money-based or­ga­ni­za­tion only leads to worse op­pres­sion by force.

Another prob­lem with his­tor­i­cal so­cial­ist states is that they un­der­mined the sense of psy­cholog­i­cal own­er­ship, al­though they could have done bet­ter (Kemp 2016).

The con­sen­sus against cen­trally planned economies is cur­rently shared not just by prac­ti­cally all economists, but also by his­to­ri­ans and so­cial sci­en­tists, and by al­most all poli­cy­mak­ers wor­ld­wide – in­clud­ing the Com­mu­nist Party of China. And the ma­jor­ity of mod­ern so­cial­ists also dis­avow cen­tral plan­ning. That be­ing said, it’s not straight­for­ward to as­sume that it won’t be re­peated. Many lead­ers of failed his­tor­i­cal so­cial­ist na­tions were avid schol­ars of Marx­ist the­ory, not mak­ing ig­no­rant mis­takes about its con­tent. So­cial­ist regimes also varied sig­nifi­cantly in their ide­olo­gies, such as the var­i­ous stages of Soviet doc­trine, Mao­ism, Juche, and other sys­tems. Many of them were even ex­plic­itly re­garded as ‘real’ so­cial­ism as dis­tinct from pre­vi­ous failures (think tank book). Yet they ended up as cen­trally-planned au­toc­ra­cies nonethe­less. More­over, a sub­stan­tial minor­ity of mod­ern leftists do defend the eco­nomic track record of these regimes, with many defend­ing Cuba in par­tic­u­lar (think tank book, much more on so­cial me­dia). This means there is a sub­stan­tial chance – we es­ti­mate about 30% –that so­cial­ism in Amer­ica would be a rep­e­ti­tion of con­ven­tional cen­tral plan­ning.

In­stead, some ad­vo­cate ParE­con, an eco­nom­i­cally rad­i­cal idea which is difficult to prop­erly eval­u­ate. How­ever it is of­ten crit­i­cized by leftists and gets lit­tle at­ten­tion these days. So­cial­ists usu­ally sup­port a more straight­for­ward in­crease in pub­lic own­er­ship and con­trol of eco­nomic de­ci­sions, ex­tend­ing the ideas of mod­ern reg­u­la­tory and welfare states and work­place democ­racy while fal­ling short of proper cen­tral plan­ning. Prob­a­bly the most no­table pro­posal for this is Sch­we­ickart (2011). We have not seen any rigor­ous, holis­tic eval­u­a­tion of such schemes, and such an eval­u­a­tion may sim­ply be im­pos­si­ble un­til they are tried. But we can look speci­fi­cally at the ma­jor com­po­nents of these vi­sions, which have been stud­ied in iso­la­tion.

One as­pect of many so­cial­ist plans is greater gov­ern­ment reg­u­la­tion of the econ­omy. But a liter­a­ture re­view (2014) col­lected 198 rele­vant em­piri­cal stud­ies pub­lished in highly se­lec­tive so­cial sci­ence jour­nals, and we add one more re­cent one (Jack­son 2017). The re­sult is that eco­nomic free­dom cor­re­sponds with good out­comes in 68% of stud­ies and bad out­comes in just 4% of stud­ies. The re­view au­thors find that this re­sult might be weak­ened by pub­li­ca­tion bias but find no ev­i­dence to in­di­cate that it would be over­turned. A smaller liter­a­ture re­view (2019) ar­gues that eco­nomic free­dom helps achieve the aims of so­cial jus­tice, which is gen­er­ally good for so­cial welfare. The think tanks which pro­duce the rank­ings of eco­nomic free­dom – mainly the Fraser In­sti­tute, but also the Her­i­tage Foun­da­tion – are con­ser­va­tive, but highly ranked (see re­ports here) and the eco­nomic free­dom rank­ings are com­monly ac­cepted in the aca­demic liter­a­ture. Such rank­ings have been crit­i­cized by the left for allegedly as­sign­ing too much free­dom to cap­i­tal­ist au­toc­ra­cies (The Guardian), but that only im­plies that they will un­der­es­ti­mate the benefits of eco­nomic free­dom. Our spe­cific eval­u­a­tions of reg­u­la­tions have found that there are some good ones (some en­vi­ron­men­tal rules, an­i­mal welfare rules, min­i­mum wages) but also bad ones (oc­cu­pa­tional li­cens­ing, zon­ing re­stric­tions, strict rent con­trols), so a sweep­ing in­crease in reg­u­la­tions across the board would likely be bad.

Another as­pect of many so­cial­ist plans is a larger gov­ern­ment with more pub­lic spend­ing. We haven’t de­ter­mined whether or not gov­ern­ment ad­minis­tra­tive agen­cies are bu­reau­crat­i­cally in­effi­cient com­pared to pri­vate com­pa­nies. How­ever, heavy pub­lic spend­ing can lead to less eco­nomic growth (Bergh and Hen­rek­son 2011, Mat­teo and Sum­merfield 2018). Of course, pub­lic spend­ing can be bet­ter tar­geted for is­sues like in­equal­ity and ex­ter­nal­ities to out­weigh eco­nomic costs, so a bal­ance must be struck. How­ever, our is­sue eval­u­a­tions have of­ten found cases where in­creas­ing gov­ern­ment spend­ing would be neu­tral or harm­ful. There­fore, sweep­ing plans for much more gov­ern­ment spend­ing seem like a poor idea.

Another as­pect of many so­cial­ist pro­grams is the re­place­ment of pri­vately owned en­ter­prises (POEs) with state-owned en­ter­prises (SOEs). Un­like tax-and-spend gov­ern­ment agen­cies, SOEs be­have in a com­mer­cial fash­ion while still be­ing di­rectly ac­countable to the gov­ern­ment. Mar­ket so­cial­ism, as prac­ticed in China, is a sys­tem where this is the main ba­sis for the econ­omy. Western so­cial­ists fre­quently re­ject it with the la­bel “state cap­i­tal­ism,” but the same rea­son­ing that the Com­mu­nist Party of China used to se­lect its poli­cies could be rec­og­nized sooner or later by a Western so­cial­ist gov­ern­ment. Two com­pre­hen­sive liter­a­ture re­views have shown that SOEs are in­fe­rior to POEs (Meg­gin­son and Net­ter 2001, Shirley and Walsh 2001). Also, two re­cent stud­ies have found that Chi­nese SOEs are in­fe­rior to POEs (Boe­ing et al 2015, Fang et al 2015). Gold­eng et al (2008) found that POEs out­performed SOEs in Nor­way in the 1990s. How­ever, some of the re­cent work ques­tions this point of view. Jakob (2017) looks at an in­ter­na­tional dataset and finds that there is no differ­ence in perfor­mance be­tween POEs and SOEs. A met­anal­y­sis by Bel et al (2010) found that pri­va­ti­za­tion of lo­cal waste and wa­ter ser­vices has no effect. Om­ran (2004) sug­gested that pri­va­ti­za­tion of Egyp­tian firms in 1994-1998 did not cre­ate sig­nifi­cant im­prove­ments. Dem­setz and Villalonga (2001) found that there is no sys­tem­atic re­la­tion­ship be­tween own­er­ship and perfor­mance in US en­ter­prises. Over­all, the ev­i­dence is rather con­flict­ing. But even if SOEs are just as good as POEs on the mar­gin, it would prob­a­bly be worse to push rad­i­cal changes to im­ple­ment many more of them. Mean­while on the the­o­ret­i­cal side, Sh­leifer and Vishny (1994) ar­gue that an econ­omy of state-run en­ter­prises merely mag­nifies the flaws of demo­cratic gov­er­nance (and Amer­i­can gov­er­nance is in­deed flawed, and will re­main so for the fore­see­able fu­ture).

Both com­mand economies and mar­ket so­cial­ism in­volve gov­ern­ment pur­suit of demo­crat­i­cally defined ob­jec­tives for the econ­omy. If we be­lieved that these ob­jec­tives would be par­tic­u­larly good, then we could spec­u­late that they might be worth the eco­nomic in­effi­cien­cies of so­cial­ism. How­ever, we can­not be op­ti­mistic about the eco­nomic goals of a so­cial­ist Amer­i­can gov­ern­ment. Our policy eval­u­a­tions in the rest of this re­port have un­cov­ered a num­ber of cases where the pop­u­lar will or lob­by­ing (in­clud­ing lob­by­ing by la­bor groups) point in poor di­rec­tions. Many of these cases are caused by struc­tural prob­lems in Amer­i­can gov­ern­ment which wouldn’t nec­es­sar­ily be fixed by a switch to so­cial­ism.

State-owned en­ter­prises re­ceive vary­ing sup­port from leftists de­pend­ing on the in­dus­try. A rel­a­tively pop­u­lar SOE pro­gram in Amer­ica would be na­tion­al­iza­tion of the fi­nance in­dus­try, out­lined by Sch­we­ickart (2011) un­der the la­bel “so­cial con­trol of in­vest­ment.” But just as in other in­dus­tries, pub­lic banks are less effi­cient than pri­vate banks (Meg­gin­son 2003). Fur­ther­more, pub­lic bank­ing in the West would cut in­vest­ment in for­eign economies, as pub­lic banks would be poli­ti­cally man­dated to sup­port pro­jects which max­i­mize em­ploy­ment for the do­mes­tic pop­u­la­tion. This would al­low more se­vere poverty in the de­vel­op­ing world.

Another ma­jor plan of many con­tem­po­rary so­cial­ists is the use of worker co­op­er­a­tives, firms which are owned and man­aged by the work­ers. Pérotin (2012, 2015) sum­ma­rizes re­search to show that worker co­op­er­a­tives have broadly pos­i­tive im­pacts. As for firms which are merely owned by work­ers, Kruse (2016) sum­ma­rized ex­ist­ing re­search to show that worker own­er­ship is mod­estly pos­i­tive for both firm perfor­mance and em­ployee welfare, though a study by Mon­teiro and Straume (2018) found in­con­clu­sive and po­ten­tially nega­tive im­pacts on firm effi­ciency in Por­tu­gal. But a sig­nifi­cant down­side of worker-owned and es­pe­cially worker-man­aged firms in the US is that they dis­cour­age out­sourc­ing to need­ier work­ers in poorer coun­tries. The 20th cen­tury’s so­cial­ist pro­grams also sug­gest that manda­tory col­lec­tiviza­tion could have very bad effects, though their prob­lems were prob­a­bly caused by state con­trol and mis­man­age­ment rather than the mere fact that they were col­lec­tives.

Leftists of­ten allege that in­sti­tu­tional bias from the cap­i­tal­ist sys­tem dis­torts the views and re­search of economists and his­to­ri­ans, so that rev­olu­tion­ary so­cial­ist pro­grams are bet­ter than they seem. But we have seen lit­tle ev­i­dence of such bias go­ing on, and it is in­con­sis­tent with two ma­jor aca­demic trends: the par­tial aca­demic pop­u­lar­ity of Marx­ist doc­trine in eco­nomics in the late 19th and early 20th cen­tury, and the wide­spread, al­most un­ques­tion­ing aca­demic ac­cep­tance of anti-cap­i­tal­ist views in cer­tain sub­fields of the hu­man­i­ties to­day. And we haven’t seen any ev­i­dence of cen­sor­ship or cen­sure of an­ti­cap­i­tal­ist views in eco­nomics de­part­ments. Con­versely, em­piri­cal anal­y­sis of so­cial­ist regimes is ac­tu­ally likely to be over-op­ti­mistic be­cause they have typ­i­cally been au­toc­ra­cies, which de­ceive their au­di­ences (Gre­gory 1990, Kor­nai 1992, Martinez 2019). Amer­i­can eco­nomics text­books sys­tem­at­i­cally over­es­ti­mated Soviet growth dur­ing the Cold War (Levy and Peart 2011). Cuba’s offi­cial statis­tics are no­to­ri­ously un­re­li­able (Salazar-Car­rillo and No­darse-León 2015, Ber­dine et al 2018).

To sum­ma­rize, gov­ern­ment con­trol of the econ­omy seems bad, heavy pub­lic spend­ing seems bad, state-owned en­ter­prise seems bad, and worker co­op­er­a­tives could be good or bad de­pend­ing on the con­text of their im­ple­men­ta­tion. How­ever, these pro­grams may have in­ter­ac­tions with each other so that a mul­ti­faceted so­cial­ist pro­gram could not be re­li­ably judged merely by tak­ing the sum of its parts. So­cial­ists usu­ally be­lieve that in­ter­ac­tions will be pos­i­tive – in other words, so­cial­ist pro­jects could perform much bet­ter than what stud­ies in­di­cate if their en­vi­ron­ment has more so­cial­ist as­pects in other ways. How­ever, we have seen no good ar­gu­ments for this view. Nega­tive in­ter­ac­tions ac­tu­ally seem more likely. The benefits of so­cial­ist pro­grams might have diminish­ing re­turns as they stack – in other words, per­haps mod­est re­forms would be suffi­cient to cap­ture the po­ten­tial benefits of so­cial­ist ideas, with ad­di­tional re­form be­ing a pointless or de­struc­tive pur­suit of un­nec­es­sary pu­rity. This is un­der­scored by the fact that the New Eco­nomic Policy, an un­usual cap­i­tal­ist pro­gram in a mostly planned econ­omy, was very suc­cess­ful. Per the Pareto Prin­ci­ple, per­haps pre­serv­ing the core 20% of cap­i­tal­ism pre­serves 80% of its benefits. Fur­ther­more, an ex­cess of so­cial­ist pro­grams can con­cen­trate too much power in the gov­ern­ment, cre­at­ing in­cen­tives and op­por­tu­ni­ties for to­tal­i­tar­i­anism and abuse as we saw in the 20th cen­tury. So while the ques­tion of in­ter­ac­tions does open up some the­o­ret­i­cal space for bet­ter so­cial­ist pro­jects to be de­vised, it doesn’t give us a good rea­son to change our point of view that a so­cial­ist sys­tem would prob­a­bly be bad.

Poli­ti­cal philos­o­phy perspectives

Co­hen (2009) ar­gues that so­cial­ism would the­o­ret­i­cally be more com­pat­i­ble with ideal prin­ci­ples like equal­ity and fair­ness, but Bren­nan (2014) demon­strates that Co­hen’s ar­gu­ment does not work. Bren­nan fur­ther ar­gues that cap­i­tal­ism is su­pe­rior in terms of the­o­ret­i­cal al­ign­ment with ideal prin­ci­ples, but Hall (2014) finds that this part of his ar­gu­ment fails. In any case, such ap­peals to ideal prin­ci­ples have lit­tle rele­vance for real eco­nomic and so­cial out­comes. The ar­gu­ment for a strong prin­ci­ple of equal­ity, where ev­ery­one ought to have equal ac­cess to a ro­bust pack­age of eco­nomic goods, has also been pushed by oth­ers (Gilabert and O’Neill 2019); how­ever, the eco­nomic ev­i­dence sug­gests that so­cial­ism fails here any­way. Sys­tems that lead to in­equal­ity might nonethe­less lead to su­pe­rior out­comes for ev­ery­one, but even leav­ing that aside, it’s plau­si­ble that ag­gre­gate welfare might be in­creased by sys­tems that leave some peo­ple worse off, so we must re­ject this strong prin­ci­ple of equal­ity. Gilabert and O’Neill seem to men­tion an idea that there is in­trin­sic im­por­tance in hav­ing eco­nomic democ­racy, but just as with poli­ti­cal democ­racy, it is only as good as its real im­pacts on welfare and this is a mat­ter for em­piri­cal study. Another proffered prin­ci­ple is the ca­pac­ity to de­velop and re­al­ize one’s own de­sired pro­jects and ac­tivi­ties (Gilabert and O’Neill 2019); how­ever this es­sen­tially boils down to eco­nomic welfare, as ob­tain­ing higher salaries, shorter work­ing hours, higher job satis­fac­tion, ear­lier re­tire­ment ages, and so on is nearly iden­ti­cal to this ca­pac­ity. Another prin­ci­ple is com­mu­nity and soli­dar­ity, with so­ciety be­ing bet­ter if peo­ple were le­gi­t­i­mately mo­ti­vated to help each other (Gilabert and O’Neill 2019). In in­trin­sic moral terms, this is false, but it holds sub­stan­tial merit in in­stru­men­tal terms. Still, it’s not clear if many forms of so­cial­ism would truly make peo­ple feel more pos­i­tively about each other. So­cial­ist move­ments them­selves seem to be driven by rather one-sided par­ti­san alle­giances, and au­thor­i­tar­ian so­cial­ist states don’t seem to have gen­er­ally had good com­mu­nal spirit, but there is plenty of po­ten­tial for things to turn out bet­ter. Utopian so­cial­ist com­mu­nal pro­jects seem to have performed well in this re­gard, but com­pa­rable in­stances of ho­moge­nous cap­i­tal­ist cul­tures have had pos­i­tive com­mu­nal spirit as well.

Similarly, Marx ar­gued that ex­ploita­tion and aliena­tion are in­her­ent to cap­i­tal­ism (Wolff 2003), but there is no good ev­i­dence show­ing that this hurts ag­gre­gate well-be­ing rel­a­tive to al­ter­na­tives. It is easy to ar­gue that the cap­i­tal­ist or­ga­ni­za­tion of alienat­ing and ex­ploita­tive la­bor (as­sum­ing Marx­ist defi­ni­tions of the terms) has led to sub­stan­tial im­prove­ments in pop­u­la­tion size and qual­ity of life over the his­tory of in­dus­trial so­ciety, and re­plac­ing it could plau­si­bly lead to in­fe­rior out­comes as we have pointed out pre­vi­ously. Numer­ous other prin­ci­pled philo­soph­i­cal ar­gu­ments have been made in fa­vor of so­cial­ism and against cap­i­tal­ism (Gilabert and O’Neill 2019) but similarly suffer from plau­si­ble coun­ter­vailing con­sid­er­a­tions and the lack of clear de­ter­mi­na­tion of real so­cioe­co­nomic out­comes such as qual­ity of life. And just as Bren­nan effec­tively re­sponded to Co­hen’s ar­gu­ments in kind with pro-cap­i­tal­ist ar­gu­ments of ideal prin­ci­ples, it seems pretty straight­for­ward to imag­ine similar re­but­tals for other, similar writ­ings. Fuer­stein (2015) mean­while ar­gues that cap­i­tal­ism weak­ens democ­racy and fur­ther shows that this weak­en­ing of democ­racy has real so­cial costs. How­ever, he does not com­pare this to pos­si­ble ways that so­cial­ist sys­tems might weaken democ­racy, and he out­lines solu­tions that are com­pat­i­ble with cap­i­tal­ism.

Note that a sur­vey of aca­demic philoso­phers found good ev­i­dence of an over­all ten­dency to dis­crim­i­nate against right-lean­ing views and in­di­vi­d­u­als (Peters et al 2019), and as the au­thors note in the be­gin­ning of the pa­per, philos­o­phy (as a gen­er­ally non-em­piri­cal dis­ci­pline) seems es­pe­cially vuln­er­a­ble to pro­duc­ing poor con­clu­sions when the re­searchers are bi­ased. While cap­i­tal­ism can of course be held as a liberal point of view, and philoso­phers are more likely to be liberal-pro­gres­sives rather than hard leftists, many of the poli­ti­cal philos­o­phy ar­gu­ments for cap­i­tal­ism are grounded in right-lean­ing prin­ci­ples (such as self-re­li­ance, eco­nomic free­dom, equal­ity of op­por­tu­nity, and op­po­si­tion to gov­ern­ment co­er­cion) which are marginal­ized by liberal-pro­gres­sives in academia. Such liberal philoso­phers fre­quently have strong egal­i­tar­ian moral be­liefs with rich con­cep­tions of restora­tive and re­dis­tribu­tive jus­tice, which means that they can only ap­peal to economists’ prag­matic ar­gu­ments for cer­tain cap­i­tal­ist poli­cies – leav­ing lit­tle that can be writ­ten in philos­o­phy jour­nals to at­tack so­cial­ism. Yet the con­ser­va­tive philo­soph­i­cal ar­gu­ments against so­cial­ism don’t seem any worse than the leftist philo­soph­i­cal ar­gu­ments for so­cial­ism. Em­piri­cally, it seems to be a pat­tern for poli­ti­cal philos­o­phy ar­gu­ments against so­cial­ism to be com­par­a­tively un­der­ex­plored in cur­rent liter­a­ture.

Peters et al did find that leftists ac­tu­ally per­ceived that they were dis­crim­i­nated against, but with­out the same sup­port­ing ev­i­dence as un­cov­ered for dis­crim­i­na­tion against the right. The far left seems un­usu­ally dis­posed to per­ceive bias and dis­crim­i­na­tion in a wide va­ri­ety of con­texts, and poli­ti­cal ex­trem­ists gen­er­ally have a less ac­cu­rate un­der­stand­ing of the be­liefs of their op­po­nents (More in Com­mon 2019), which casts more doubt on this sur­vey re­sult. That be­ing said, it is plau­si­ble that far-left views do re­ceive real dis­crim­i­na­tion. So over­all, there are grounds for just a mod­er­ate pre­sump­tion that poli­ti­cal philos­o­phy has left­ward bias here. But in gen­eral, the poli­ti­cal philos­o­phy per­spec­tives do not provide sig­nifi­cant rea­son to change our point of view.

Prac­ti­cal dimensions

So­cial­ist change would likely come in the form of vi­o­lent rev­olu­tion, which would have a num­ber of ad­di­tional do­mes­tic and in­ter­na­tional costs. His­tor­i­cal at­tempts at com­mu­nist sys­tems were never in­tro­duced as a con­se­quence of the party win­ning an elec­tion, it has only hap­pened through force (Kemp 2016). Amer­i­can leftists have in­creas­ingly uti­lized or em­braced vi­o­lent tac­tics in the last sev­eral years, a pat­tern which has caused lit­tle di­rect harm but in some forms (an­tifas­cist, anti-Trump) gar­ners a rel­a­tively high level of sym­pa­thy from some me­dia, academia, poli­ti­cal ac­tors and broader pub­lic, thus cre­at­ing a sig­nifi­cant pos­si­bil­ity for it to con­tinue and grow in the event of a mass so­cial­ist move­ment. The no­tion of vi­o­lent rev­olu­tion is com­monly en­dorsed in leftist cir­cles, with some call­ing it in­evitable. The So­cial­ist Rifle As­so­ci­a­tion in the US has a cou­ple thou­sand mem­bers, al­though its be­hav­ior and rhetoric have been fine from what we have seen. Even if so­cial­ism were de­sir­able, it may not be good enough to out­weigh the risks of vi­o­lence. And even if so­cial­ism were de­sir­able enough to out­weigh the costs of vi­o­lence, the event of a failed vi­o­lent rev­olu­tion would still be a clearly bad thing. This gives an ex­tra rea­son to pre­fer main­te­nance of the ex­ist­ing eco­nomic sys­tem and avoid in­sur­rec­tions in the first place.

There is at least some value of in­for­ma­tion in test­ing a bet­ter model of so­cial­ism, if a good path­way to run­ning a good test can be iden­ti­fied. No one has at­tempted the most re­cent, re­fined so­cial­ist plans. If a na­tion demon­strated that they can work well, then many other na­tions could im­prove their own poli­cies ac­cord­ingly. If a na­tion tried and failed, then they could even­tu­ally re­turn to cap­i­tal­ism (though his­tor­i­cally this has been a slow and painful pro­cess) and their ex­pe­rience would in­form peo­ple in other coun­tries to re­frain from pur­su­ing the mat­ter. This benefit is sub­stan­tial and should be taken se­ri­ously even in the face of gen­eral skep­ti­cism about so­cial­ism. How­ever, it is coun­tered by the risk that a so­cial­ist move­ment or gov­ern­ment would re­duce the prospects for ex­per­i­men­ta­tion with differ­ent kinds of eco­nomic pro­pos­als. Other ideas for ma­jor so­cioe­co­nomic re­form such as com­mu­nal­ism, Ge­or­gism and liberal rad­i­cal­ism fall upon hos­tile, deaf or at least merely aca­demic ears when so­cial­ism is as­sumed to be the de­fault rem­edy for the per­sis­tent ills of cap­i­tal­ism. Both Ge­or­gism and liberal rad­i­cal­ism ap­pear bet­ter than so­cial­ism (though they are not nec­es­sar­ily mu­tu­ally ex­clu­sive), but nei­ther has been prop­erly tested even once.

Ex­per­i­men­tal value would be more im­por­tant if cap­i­tal­ism were lead­ing to a great crisis. If busi­ness as usual were to lead to a ma­jor risk of catas­trophic out­comes like en­trenched cap­i­tal­ist aris­toc­racy, global fas­cism, or dev­as­tat­ing cli­mate change, then test­ing some­thing – any­thing – to avert the dis­aster would have more ur­gency. How­ever, such pes­simistic pre­dic­tions gen­er­ally seem false. Ob­jec­tive mea­sure­ment of var­i­ous in­di­ca­tors of the health of so­ciety shows that they are mostly im­prov­ing (Our World in Data).

Look­ing at the gen­eral de­gree of un­cer­tainty in the is­sue, and tak­ing ex­per­i­men­tal value into ac­count, we are un­sure about whether one could sketch out a fea­si­ble so­cial­ist pro­posal for the US that would be worth do­ing over the cur­rent trend of cap­i­tal­ist poli­cies across the ad­vanced world. How­ever, that doesn’t mean that the ac­tual re­sults of a so­cial­ist move­ment would meet this stan­dard. In fact, there are rea­sons to be speci­fi­cally pes­simistic about the re­sults of a so­cial­ist move­ment. Trends among many cur­rent Western leftists – over­con­fi­dence in their point of view, dis­in­ter­est in policy plan­ning, eco­nomic de­nial­ism and folk-eco­nomic be­liefs, apolo­gia for to­tal­i­tar­ian so­cial­ism (Quillette, think tank book, much more on so­cial me­dia), de­hu­man­iza­tion and vi­o­lence to­wards poli­ti­cal op­po­nents, cen­sor­ship and cen­sure of in­ter­nal dis­sent, in­ter­nal sec­tar­i­anism, re­jec­tion of sci­en­tific ob­jec­tivity and ra­tio­nal­ity, and vin­dic­tive at­ti­tudes on so­cial jus­tice poli­tics (Amer­i­can Con­ser­va­tive) – in­crease the chances of re­peat­ing the failure modes of 20th cen­tury so­cial­ist pro­grams. Th­ese trends might be con­sid­ered rel­a­tively be­nign in the cur­rent con­text, where leftists are a minor­ity in a so­ciety with sta­ble in­sti­tu­tions. But the trends be­come more dan­ger­ous if they are preva­lent in a rev­olu­tion­ary fac­tion which can as­sert its own in­sti­tu­tions and lead­er­ship. It’s worth not­ing that Lenin’s vi­sion for the USSR ac­tu­ally in­volved a fair amount of ex­per­i­men­tal think­ing, open-mind­ed­ness and re­spect for ci­vil­ity, be­fore Stalin took power and over­turned these early ideas (Žižek 2011 ap­pendix); a mostly de­cent move­ment can be hi­jacked by un­sa­vory el­e­ments if the pro­cess of rad­i­cal change has un­done the norms and in­sti­tu­tions re­spon­si­ble for pro­tect­ing the gov­ern­ment from such a de­vel­op­ment.

More­over, em­pow­er­ing a so­cial­ist move­ment could have more effects be­sides so­cial­ism it­self. So­cial­ists may lev­er­age their power to in­stall a suite of other pro­grams, which might be in­ferred from the con­tent of the So­cial­ism 2019 con­fer­ence: trans­fem­i­nism, black liber­a­tion, open bor­ders, abor­tion ac­cess, anti-Zion­ism, strong en­vi­ron­men­tal­ism, and weak­en­ing (if not abol­ish­ing) the po­lice and mil­i­tary. So­cial­ists have also fre­quently been hos­tile to philan­thropy, be­liev­ing it to be un­nec­es­sary in the con­text of a so­cial­ist state (Twit­ter), villify­ing wealthy philan­thropists and re­peat­ing dis­cred­ited ar­gu­ments against Effec­tive Altru­ism. The over­all de­sir­a­bil­ity of this broad suite of sec­ondary am­bi­tions is un­clear, which adds an­other layer of un­cer­tainty on the mat­ter. Note how­ever that some of these ten­den­cies among so­cial­ist elites are likely to be diluted by com­mon sen­ti­ments in a mass so­cial­ist move­ment.

So­cial­ist be­liefs also lead peo­ple to take po­si­tions of in­differ­ence or out­right hos­tility to more rea­son­able im­me­di­ate re­forms as long as we re­main in a cap­i­tal­ist econ­omy. For in­stance, Ka­mala Har­ris’ pro­posal to try new af­ter-school care pro­grams which would re­lieve the child care bur­den off work­ing fam­i­lies was met with hos­tility by a num­ber of so­cial­ists who saw it as con­trary to their vi­sion of elimi­nat­ing long work­ing days en­tirely. How­ever, there is a con­verse prob­lem where pop­u­lar an­tipa­thy to­wards so­cial­ism causes con­ser­va­tives to op­pose rea­son­able plans like uni­ver­sal health­care.


Over­all, there is good but not de­ci­sive ev­i­dence against so­cial­ism gen­er­ally speak­ing. It’s an open ques­tion whether one could de­sign a par­tic­u­lar form of so­cial­ism – ei­ther one which lev­er­ages worker co­op­er­a­tives while main­tain­ing com­pet­i­tive mar­kets and a rea­son­ably sized gov­ern­ment, or one which has some so­phis­ti­cated and well-in­cen­tivized form of cen­tral plan­ning – that would be worth test­ing in place of the cur­rent cap­i­tal­ist sys­tem. How­ever, for those who are in­ter­ested in rad­i­cal change, we think it is fair to say that there are prob­a­bly more worth­while pos­si­bil­ities be­sides so­cial­ism.

In ad­di­tion, the re­al­ity of a so­cial­ist sys­tem if Amer­i­cans de­cide to im­ple­ment one will very prob­a­bly be worse than ideal so­cial­ism and worse than the sta­tus quo. And there is no guaran­tee that so­cial­ist change would be a peace­ful pro­cess. There­fore, we should pre­fer that Amer­i­cans have more faith in cap­i­tal­ism and re­sis­tance to so­cial­ism.

This find­ing is too weak and un­cer­tain to be jus­tified as a main cause pri­or­ity. How­ever, it still has in­ter­est­ing im­pli­ca­tions for Effec­tive Altru­ism. Numer­ous writ­ers have alleged that char­ity is worse than it seems be­cause it could re­in­force cap­i­tal­ism, and ar­gued that EAs have a duty of jus­tice to pro­mote new eco­nomic sys­tems that could help the de­vel­op­ing world es­cape poverty. How­ever, the find­ing here means that this ar­gu­ment must be re­versed. Char­ity is ex­tra good be­cause it re­in­forces cap­i­tal­ism, and EAs have a duty of jus­tice to re­in­force (and re­fine) cap­i­tal­ism in or­der to help the de­vel­op­ing world. Of course nei­ther of these ar­gu­ments ac­tu­ally make sense—the num­ber of peo­ple whose poli­ti­cal at­ti­tudes are ac­tu­ally changed by philan­thropy is neg­ligible as far as any­one can tell, and the “duty of jus­tice” is a moral false­hood. But as far as philo­soph­i­cal de­bates are con­cerned, a good deal of aca­demic or­tho­doxy will have to be re­vised in the light of this new idea that cap­i­tal­ist sys­tems are usu­ally su­pe­rior to so­cial­ist sys­tems.